From: @ling_qiao_min
Reviewed-by: @zhang_xue_tong,@hangangqiang
Signed-off-by: @zhang_xue_tong

From: @pengyongrong
Reviewed-by: 
Signed-off-by:

Cast the float16-input to float32 before ReduceSum, and cast back to float16 after ReduceSum.
If the op after this ReduceSum is a casting from float16 to float32, then it can be eliminated.

Fix can not collect input data when batch size is 1 and total step
number is 1

Fixed spelling errors

From: @sishuikang
Reviewed-by: @zhanghaibo5,@zhang_xue_tong
Signed-off-by: @zhanghaibo5

From: @jjfeing
Reviewed-by: @zhoufeng54,@kisnwang
Signed-off-by: @kisnwang

From: @jianghui58
Reviewed-by: 
Signed-off-by:

From: @probiotics_53
Reviewed-by: @hangangqiang,@zhang_xue_tong
Signed-off-by: @hangangqiang

From: @gaoyong10
Reviewed-by: 
Signed-off-by:

From: @c_34
Reviewed-by: @liangchenghui
Signed-off-by:

From: @hfarahat
Reviewed-by: 
Signed-off-by:

From: @yanglf1121
Reviewed-by: @wuxuejian,@kisnwang
Signed-off-by: @wuxuejian

From: @mengyuanli
Reviewed-by: @hangangqiang,@zhanghaibo5
Signed-off-by: @hangangqiang

From: @liubuyu
Reviewed-by: @zhoufeng54,@jjfeing
Signed-off-by: @jjfeing

From: @yangruoqi713
Reviewed-by: @hangangqiang
Signed-off-by:

From: @yeyunpeng2020
Reviewed-by: 
Signed-off-by:

From: @yangruoqi713
Reviewed-by: @zhang_xue_tong,@hangangqiang
Signed-off-by: @zhang_xue_tong

From: @chenzupeng
Reviewed-by: 
Signed-off-by:

From: @sishuikang
Reviewed-by: @zhanghaibo5,@zhang_xue_tong
Signed-off-by: @zhanghaibo5

From: @sishuikang
Reviewed-by: 
Signed-off-by:

From: @anancds
Reviewed-by: @xsmq,@zh_qh
Signed-off-by: @zh_qh

From: @xukailun_1
Reviewed-by: @wuxuejian,@liangchenghui
Signed-off-by: @wuxuejian

From: @jiangzg001
Reviewed-by: @liangchenghui,@wuxuejian
Signed-off-by: @wuxuejian

fix some bugs